Oracle, Which Promised To Protect TikTok User Data From China, Helps Chinese Law Enforcement Snarf Through Lots Of Private Data

from the oh-look-at-that dept

As you'll recall, last summer there was a whole performative nonsense thing with then President Trump declaring TikTok to be a national security threat (just shortly after some kids on TikTok made him look silly by reserving a million tickets to a Trump rally they never intended to attend). Trump and his cronies insisted that TikTok owner ByteDance had to sell the US operations of TikTok to an American firm. The whole rationale about this was the claim -- unsupported by any direct evidence -- that TikTok was a privacy risk, because it was owned by a firm based in Beijing, and that firm likely had connections to the Chinese government (as do basically all large Chinese firms). But how was that privacy risk any worse than pretty much any other company? No one ever seemed to be able to say.

Eventually, after Trump blocked both Microsoft and Walmart from doing the deal, he "approved" a non-sale, but "hosting" deal with Oracle, whose founder/chair, Larry Ellison, and CEO, Safra Catz, were both big Trump supporters. It quickly came out that TikTok's investors deliberately went hunting for a company that they knew Trump liked, and that's why they asked Oracle.

But, part of the announcement of the "deal" was that Oracle would make sure that US TikTok users had their data protected, and that Oracle would keep that data outside the hands of the Chinese government. That seemed somewhat rich, considering that Oracle's initial rise to being a tech giant was built almost entirely on its close connections to the US government, and specifically the intelligence agencies. But it's become even more rich now that the Intercept reports that Oracle actually has a lucrative business helping repressive law enforcement in China do surveillance work. The long story is absolutely full of totally shocking -- but somehow not surprising -- details. It starts off by noting that Oracle hosted a presentation on its own website, literally describing how it helped police in Liaoning province better sort through all of the surveillance data they collected:

Police in China's Liaoning province were sitting on mounds of data collected through invasive means: financial records, travel information, vehicle registrations, social media, and surveillance camera footage. To make sense of it all, they needed sophisticated analytic software. Enter American business computing giant Oracle, whose products could find relevant data in the police department’s disparate feeds and merge it with information from ongoing investigations.

So explained a China-based Oracle engineer at a developer conference at the company’s California headquarters in 2018. Slides from the presentation, hosted on Oracle’s website, begin with a “case outline” listing four Oracle “product[s] used” by Liaoning police to “do criminal analysis and prediction.” One slide shows Oracle software enabling Liaoning police to create network graphs based on hotel registrations and track down anyone who might be linked to a given suspect. Another shows the software being used to build a police dashboard and create “security case heat map[s].” Apparent pictures of the software interface show a blurred face and various Chinese names. The concluding slide states that the software helped police, whose datasets had been “incomprehensible,” more easily “trace the key people/objects/events” and “identify potential suspect[s]” — which in China often means dissidents.

And, yes, if you're wondering, apparently Oracle is helping police in Xingjiang, where there has been ongoing genocide happening against Uyghur Muslims.

In marketing materials, Oracle said that its software could help police leverage information from online comments, investigation records, hotel registrations, license plate information, DNA databases, and images for facial recognition. Oracle presentations even suggested that police could use its products to combine social media activity with dedicated Chinese government databases tracking drug users and people in the entertainment industry, a group that includes sex workers. Oracle employees also promoted company technology for China’s “Police Cloud,” a big data platform implemented as part of the emerging surveillance state.

Several Oracle materials imply that the company has gone substantially further than marketing to Chinese police, which operate as part of the country’s Ministry of Public Security: One presentation detailing Oracle’s database and data security products contains a slide titled “Oracle and the national defense industry.” That title is followed by a list of multiple Chinese military entities, including the People’s Liberation Army, China National Nuclear Corporation, and China Aerospace Science and Technology Corporation. Defense entities are also the apparent target for two additional Oracle Chinese-language presentations, the most recent of which is dated 2015, and for events called the “People’s Armed Police Force–Oracle Cloud Computing Exchange Forum” and the “Oracle Xi’an Aviation and National Defense Industry Informatization Seminar” listed in Chinese on Oracle’s site.

Yikes.

So the whole blasted pitch about taking control over TikTok was about keeping the data away from the Chinese, while at the very same time, the same company is helping Chinese law enforcement scan through tons of surveillance data to better suppress its people?

If you're wondering how Oracle responded to the report, it had a spokesperson claim that the examples in these documents were "theoretical" pitch decks to show how the technology could be used, not how they were being used. But as the article notes, that appears to not be true, and the presentation makes it clear that it's talking about an actual case study, and even points out that the police supplied their own data, which they then analyzed with Oracle's technology.

The article is quite long, but turns up a shocking number of smoking guns:

Some of the Chinese-language presentations on Oracle’s site are labeled “CONFIDENTIAL,” despite being publicly available. It is easy to see why someone might have wanted to keep them hidden. Taken together, they show an extreme willingness to aid in the construction of the surveillance state. One Chinese-language presentation, for example, promotes “Oracle’s recommendation: a more complete platform to meet the needs of public security big data processing.”

[....]

Another pitch depicts a broad array of sensitive citizen data being converted into ones and zeros, including DNA, mental illness records, and other medical information. Still other documents from China boast that Oracle technology can help police trawl internet activity to “analyze potential suspected criminal behavior among hundreds of millions of netizens,” capture license plate data from “tens of thousands of cameras,” and analyze call records to build out criminal networks, then link them to fingerprint and facial recognition images.

Of course, it's not clear if the Oracle TikTok deal will ever happen. It seems that no one was much interested in it beyond the headlines, and the Biden administration doesn't seem keen on enforcing the executive order that created the need for Oracle to step in. And, honestly, after reading this report, perhaps that's much better for the privacy of Americans using TikTok.

Filed Under: china, law enforcement, police, privacy
Companies: bytedance, oracle, tiktok

Indian Government Requires Educational Establishments To Obtain Its Approval For The Subject Matter And Participants Of International Online Conferences And Seminars

from the hardly-an-edifying-sight dept

It would be something of an understatement to say that the COVID-19 pandemic has had a big effect on our lives. One sector where people have had to re-invent themselves is the academic world. Core in-person activities like lectures, seminars, and conferences have been forced to move online. One advantage of a shift to virtual gatherings is that people can participate from around the world. However, for governments, that's less a feature than a bug, since it means they have less control over who is taking part, and what they are saying. In response to this development, the Ministry of Education in India has issued "Revised Guidelines for holding online/virtual Conferences, Seminars, Training, etc." (pdf). An opinion piece in The Indian Express calls it the "biggest attack in the history of independent India on the autonomy of our universities":

When it is fully enforced -- and let there be no doubts over the government's resolve to be iron-handed when it comes to restricting people's democratic rights -- India will find itself in the company of dictatorial regimes around the world that despise liberty of thought and muzzle freedom of expression in their institutions of higher learning.

The new guidelines apply to all publicly funded higher education establishments. The key requirement is for international online conferences and seminars to avoid politically sensitive topics, specifically any related to problems along India's borders. Chief among these are disputes between India and China over borders in the north-east of India, which has recently seen skirmishes between the Indian and Chinese armies, and in Kashmir. Another ban, vague in the extreme, concerns "any other issues which are clearly/purely related to India's internal matters". As well as obtaining approval for the topics of planned online meetings, educational establishments must also submit a list of participants to be vetted. And once an approved conference or seminar has taken place, a link to the event must be provided. As The Indian Express column points out, these new restrictions are likely to hit Indian universities particularly hard:

Unlike their western counterparts, they are severely under-funded. They can neither organise many international conferences, nor send their faculty to participate in such events abroad. The recent boom in webinars has, hence, come as a big boon to them. It saves travel and hospitality costs and also overcomes the hassles of getting visas for invitees from "unfriendly" countries. Moreover, such events can be easily, and more frequently, organised even by institutions in rural and remote areas. Disturbingly, the government wants to curtail these major benefits of the digital revolution to millions of our teachers, students and scientists.

The Indian government's desire to control what is said, and by whom, is likely to harm the spread of knowledge in a country that was just beginning to enjoy one of the few benefits of the pandemic: easier access to international academic gatherings.

Follow me @glynmoody on Twitter, Diaspora, or Mastodon.

Filed Under: academic freedom, china, conferences, free speech, india, pandemic

Steam Becomes Available In China, Offers 53 Whole Games To Customers

from the ultimate-curation dept

There is no shortage of critiques for Valve's online PC game store, Steam. That's to be expected, frankly, given how big the platform is. Still, on the ground with individual gamers, one of the most common complaints you hear will be that the sheer volume of games on Steam is somewhat paralyzing for customers deciding where to spend their money. Steam tried to combat this for years with its Steam Curators program, where gamers put their trust in curators to pare down game search results. It never really worked, though, as the program encountered the same issue as the game: the sheer volume of curators.

And so nothing really got solved. Except for in China, it seems, where Steam recently launched with a grand total of 53 whole games available to buyers.

Steam is now available in China, and if you thought that would give Chinese gamers instant access to the weird, wonderful and sometimes deeply offensive depths of the service’s catalogue, well lol, no, of course it doesn’t.

At time of posting this custom, localised version of Valve’s shopfront only has 53 games available, with the main ones being Counter-Strike and DOTA 2. That’s it. If you visit the store and click on “all games” you can see everything in a single screenshot.

You already know why this is happening. In order to be available in China, games have to go through an approval process with the Chinese government. Given the insanely strict rules for approval, few publishers even try to get approval. That, plus the related apathy for entering the market, gets you 53 games on Chinese Steam.

In addition, the forums on Steam are blocked as well, ostensibly to keep any speech Beijing would disapprove of from appearing on the site. Oddly, user reviews are available, however.

Honestly, it's enough to make you wonder why this release was worth it for Steam at all. Put another way, if gaining approval from the Chinese government to release a video game is so difficult and/or arduous to keep most game publishers away, why would a game storefront be any different?

Filed Under: china, steam, video games
Companies: valve

Trump And Oracle's Dumb TikTok Cronyism Falls Apart

from the end-of-the-line dept

Remember when America spent a year and a half hyperventilating about a Chinese teen dancing app instead of securing American infrastructure from Russian hackers or other threats? Remember when a bunch of GOP officials with a long track record of not caring whatsoever about consumer privacy or internet security exploited xenophobic fears about the app to land political allies Oracle and Walmart a major windfall? Remember when 90% of the press couldn't be bothered to inform readers this was all performative cronyism by an unqualified nitwit? Good times.

This morning the Wall Street Journal announced that the much hyped deal to sell ByteDance-owned TikTok to Oracle and Walmart is looking unsurprisingly dead in the wake of previous legal challenges and Trump's election loss. Instead, the government appears poised to do what made sense from the start: focus on the broader problem of lax privacy and dodgy security standards across the board in telecom/adtech/tech, instead of singling out a teen dancing app:

"We plan to develop a comprehensive approach to securing U.S. data that addresses the full range of threats we face,” National Security Council spokeswoman Emily Horne said. “This includes the risk posed by Chinese apps and other software that operate in the U.S. In the coming months, we expect to review specific cases in light of a comprehensive understanding of the risks we face."

Again, this whole effort was about little more than Trump trying to gain leverage in a ham-fisted and unproductive trade war, while using the power of his office to hand over a successful business to his political buddies at Oracle and Walmart. Yet somehow, they managed to get the majority of the US technology press to believe this was some genuine, good faith effort to shore up consumer privacy. There were a few exceptions to this rule in press and policy circles, but they were the exception, not the norm:

Of course it was never really about security and privacy, since there's a parade of other more pressing issues Trump and the GOP clearly didn't care about. Trumpland opposed absolutely anything even resembling consumer privacy rules and standards for the internet era. It opposed funding improved election security. It clearly failed to secure US networks from hackers, and did little to nothing to shore up very real US security vulnerabilities be it the SS7 flaw or the dumpster fire that is the internet of broken things. They didn't genuinely care about internet privacy and security. They just didn't.

As such this wasn't just a failure for Trump, it was an immense failure for the US technology press, which successfully parroted Trump's claims that they were just really worried about consumer data privacy, despite zero evidence on any other front that this was actually true. While the Journal tries to feebly suggest that a deal could still be struck to sell TikTok, that doesn't seem likely. Instead the saga appears to be dying with a sad, quiet whimper, with the only winners being those who gobbled up the millions in billable legal hours to pursue this idiotic tech policy brain fart.

Filed Under: china, donald trump, larry ellison, performative nonsense, political theater, privacy, security theater
Companies: oracle, tiktok, walmart

Huawei Attempts To Rebuild Trust By Using... Fake Twitter Telecom Experts

from the not-helping dept

So we've noted that a lot of the accusations that Huawei spies on Americans on behalf of the US government are lacking in the evidence department. The company's been on the receiving end of a global blocklist based on accusations that have never actually been proven publicly, levied by a country (the United States) with a long, long history of doing exactly what it accuses Huawei of doing. While scrutiny of Chinese gear is certainly warranted, at the same time it's a rather idiotic rabbit hole filled with xenophobic politicians being played by US companies that like to drum up NatSec hysteria for political advantage.

That of course doesn't mean Huawei, like so many telecom giants, isn't a terrible company that routinely makes unethical decisions. Like when the company was busted helping several different African governments to spy on political opponents and journalists. Or when it was shown to be eagerly helping China build facial identification systems to quickly identify Uighur Muslims. Or, more recently, when the company was busted creating entirely fake people in a ham-fisted bid to try and boost its reputation.

Last week, the company was busted for creating at least fourteen different Twitter accounts pretending to be respected telecommunications experts, writers, and assorted academics. Their goal, to cast doubt on Belgian legislation attempting to limit "high risk" vendors from building the nation's 5G networks. The bogus, pro-Huawei accounts used computer generated profile pictures. Their Tweets were then amplified by official, and real, Huawei executives:

"Kevin Liu, Huawei’s president for public affairs and communications in Western Europe, who has a verified Twitter account with 1.1 million followers, shared 60 posts from the fake accounts over three weeks in December, according to Graphika. Huawei’s official account in Europe, with more than five million followers, did so 47 times."

To be clear, creating fake people to generate fake support for various policy proposals certainly isn't unique. US companies do it fairly routinely to generate fake buzz and shape the discourse on social media platforms. Fake people were used to applaud the FCC's repeal of net neutrality at the telecom industry's behest just a few years back. Chinese Pro-Trump media outlet, The Epoch Times, was also busted using bogus people to push conspiracy theories the last few years.

It's quite the blossoming little business sector. One recent report out of Oxford identified 63 examples in which public relations firms were involved in online disinformation operations in 2020 for government or corporate clients. In this case, Twitter took down the fourteen fake accounts immediately after being notified by the New York Times. Graphika, the company that first unearthed the astroturf effort, connected the dots after it noticed that several accounts used in previous campaigns began sharing pro-Huawei arguments and attacking Belgium's regulatory approach.

A common practice or not, it's monumentally idiotic to think that the best way to show you're trustworthy as an international telecom vendor is to create fake people to make your case for you. Then, apparently, assume you were being so clever that nobody would ever discover the charade. But again, much like spying on your own citizens at the behalf of government, it's something companies like AT&T also engage in, which undermines any serious attempt by the US to hold the moral high ground.

Filed Under: china, fake experts, trust
Companies: huawei

Trump's Facts-Optional Assault On Chinese Tech Continues With Blocking Of Xiaomi

from the glass-houses dept

All of these statements can be true:

• The Chinese government engages in all manner of dangerous, unethical behavior and Chinese companies should be carefully monitored for threats to national security, facing productive penalties when public evidence is provided.
• The US government engages in all manner of dangerous, unethical behavior and US companies should be carefully monitored for threats to national security, facing productive penalties when public evidence is provided.
• Some of the US hyperventilation over Chinese tech and telecom is xenophobic nonsense exploited by US companies for financial gain.

Throughout the Trump administration's assault on Chinese companies, transparent evidence of wrongdoing has been an afterthought. There was little to no evidence provided by the administration to justify its year-long assault on TikTok, which arguably distracted government leaders from far more pressing threats to national security. Similarly, while blocklisted network hardware vendor Huawei engages in some atrocious behavior around the world, evidence that the company directly spies on US citizens at the behest of China (the core accusation) has also never been publicly proven, despite ample hyperventilation.

Fast forward to last week, when the Trump administration announced it had blocklisted Chinese smartphone maker Xiaomi for its alleged ties to the Chinese military:

"The Defense Department said that Xiaomi, along with eight other firms it added to its blacklist, was working to support China’s People’s Liberation Army by giving it access to its technology. In a statement, the Defense Department accused Xiaomi of being part of an orchestrated campaign to support the "modernization goals of the People's Liberation Army by ensuring its access to advanced technology and expertise."

The problem, again, is that there's no evidence of this, and when the shoe is on the other foot, US companies get understandably angry. Xiaomi, which recently passed Apple to become the third largest smartphone manufacturer in the world, has denied any connection to the Chinese military, and the Trump administration has provided no evidence that one exists. A not insubstantial portion of these efforts are driven by xenophobia and greed (in many instances US companies just don't want to compete with cheaper kit), yet there's an over-arching pretense in the press that this somehow isn't the case.

The other problem we like to downplay is the US' blistering hypocrisy on this subject. US telecom giant AT&T is fused so tightly to our global intelligence efforts, you'd have a hard time physically determining where AT&T ends and the NSA begins. The US had no problem breaking into Huawei to install its own backdoors, or stealing network hardware in transit to install backdoors, or, you know, spying on absolutely everybody, including allies, all the time, for several generations, with zero respect for boundaries or the rule of law. We provided IT support to a continent of brutal fascists during Operation Condor, for fuck's sake.

If the US wants to be taken seriously on this subject, it has a hell of a lot of work to do, starting with leading by example. And again, if there's direct evidence that a Chinese company has been spying on Americans, then provide it and let's have a transparent conversation about it. But most coverage and discussion of Trump's war on Chinese tech involves pretending a lack of public evidence is okay, pretending the US' dodgy history on this very subject doesn't exist, pretending that a lot of these efforts aren't driven by bigotry or simple greed, and pretending that the Trump administration has any idea whatsoever what it's doing.

Filed Under: china, donald trump
Companies: xiaomi

The Beijing Arrests In Hong Kong Expand As America Loses Its Ability To Credibly Respond

from the democracy-on-decline dept

While we've been discussing the troubling actions the Chinese government has undertaken against Hong Kong for some time now, the movements are starting to reach a rapid pace. It was only in July that Beijing unveiled a sparkling new and impossibly vague "national security" law that was perfectly tuned to make any pro-democracy talk or demonstrations in Hong Kong offenses that would result in lengthy prison time. This was the result of months of protests that served to embarrass a Communist government that prides itself on full control. Again, that was only June. In August, the arrests of media members who were pro-democracy began, followed by the ouster and then arrest of pro-democracy lawmakers in November.

And while one can imagine the Chinese government doing a bit of breath-holding in December as the American political system buried itself in bullshit thanks to one man's inability to admit defeat, it appears January was the start date to re-kick off the attacks on Hong Kong. In the early morning hours of Wednesday, 53 Hong Kong citizens were arrested under the national security law, all accused of, well...

By late morning on Wednesday, at least 53 Hong Kong residents — former lawmakers, activists and an American lawyer among them — had been detained under Beijing’s new national security law, and their offices and homes raided. Accused of subversion, they face up to life in prison for holding a primary vote last year ahead of legislative elections that were ultimately postponed and which many of them were barred from contesting.

The raids, which involved more than 1,000 officers, marked the most dramatic onslaught in Chinese leader Xi Jinping’s transformation of once-freewheeling Hong Kong into a city gripped by fear under authoritarian rule.

This was predicted by many, me included. With Trump never all that concerned with human rights in China and now all the more distracted with a combination of rounds of golf to be played and living in some aggrieved fantasy land, the Chinese government is quite clearly going to take every authoritarian action it can now, before the new administration is in place. This should also serve as a lovely indication of just how dumb the claims are that Biden and Harris are somehow compromised by China. Were that true, China would be holding off until January 20th, and yet they are not.

The Hong Kong authorities are, as ever, playing along with their puppet-masters.

John Lee, Hong Kong’s secretary for security, lauded the police operation as “necessary” and accused those involved in the primary of attempting to paralyze the government by seeking to win a majority in the legislature.

I'll give you a moment to reread that last bit again. Those arrested are accused of trying to "paralyze" Hong Kong's government by winning elections. The argument, in other words, is that pro-democracy advocates are too dangerous to be allowed to be elected, so they are being forcibly removed from society via aggressive actions instead.

It's almost poetic, then, that these arrests came on the same day that Donald Trump and his enablers fostered and promoted an insurrection by his supporters at the Capitol, all to achieve precisely the same goal. That it failed is good, but of no material importance when it comes to America's current and future standing to apply pressure on China to cease its actions in Hong Kong. You can be damned sure that Chinese state television will be playing the videos and images of the chaos that occurred in DC on Wednesday on loop for its citizens as a way to rebut any words Joe Biden might have on the topic of Hong Kong.

And so, as authoritarian regimes light fires around the world, it's worth pausing and reflecting on just how much was lost on Wednesday in America. Our dignity. Our peace. Our moral authority. Our standing.

Filed Under: arrests, china, democracy, donald trump, free speech, hong kong

Hong Kong Court Revokes Bail For Jimmy Lai After Deciding It Didn't Interpret Vague National Security Law Vaguely Enough

from the criminalizing-impolite-disagreement dept

The Chinese government has been showing its impatience over its impending takeover of Hong Kong. China agreed to allow Hong Kong to run under its own government until 2047, but the last couple of years have seen the Chinese government indicating its willingness to perpetually violate this agreement with the region.

As pro-democracy protests continue to rage against the Chinese machine, the Chinese government has begun forcing its will on Hong Kong residents. This has been greatly aided by complicit Hong Kong government legislators, who have basically agreed to all the Chinese government's demands. The latest attempt to undermine the will of the people came packaged in a "national security" law -- one that outlawed demonstrations against the Chinese government's early and uninvited interloping, threatening dissenters with a lifetime of imprisonment.

Since then, the Chinese government (again with the assistance of Hong Kong's supposedly-independent government) has been arresting and jailing prominent critics, along with dozens of other vocal protesters. One of the most famous arrestees is Jimmy Lai, a vociferously pro-democracy media tycoon -- one with the power and reach to do serious damage to the Chinese government's unwanted advances.

Lai was arrested under the new national security law last August. The Chinese government -- via its Hong Kong mouthpiece -- claimed Lai's pro-democracy agitation was an illegal "collusion" with "foreign governments." This vague assertion likely referred to Lai's visits with world leaders -- visits in which he expressed his displeasure with the Chinese government.

In late December, Lai was granted bail by a Hong Kong court, reversing its earlier denial. It came with very tight strings attached. Lai was placed on house arrest and, more importantly, forbidden from doing anything that might make the two governments engaged in his prosecution look bad.

[Lai was] banned from meeting officials from foreign governments, attending or hosting media interviews or programmes, publishing articles in any media, and posting messages or comments on social media, including Twitter.

He was also hit with a $10 million HKD fine ($1.3 million USD) and ordered to surrender all travel documents.

The government apparently felt this punishment wasn't heinous enough. Or, at the very least, unlikely to silence the outspoken government critic. It went back to the court and got the reversal reversed, putting Lai back behind bars. This time around, the court bought the government's national security arguments, relying on unexplored areas of the brand new law to deprive Lai of his very limited freedom.

In its determination, the court held that it was “reasonably arguable … that the learned judge may have erred in his construction or application” of Article 42(2) of the national security law, under which bail can only be granted to a criminal defendant if a judge has sufficient grounds to believe that the defendant will not commit acts endangering national security.

The new law apparently makes speaking out against the Chinese and Hong Kong governments a national security issue, even if it's nothing more than speaking out against the Chinese government's violation of an agreement it struck with the UK back in 1997. This is the problem with national security laws, no matter where they're deployed. A little imagination is all it takes to turn dissent into a threat and outspoken criticism into a criminal act. That's why there's so many of these laws in countries that thrive on punishing dissent. And that's why countries that are considered more "free" find them handy to have around, just in case.

Filed Under: authoritarianism, china, democracy, hong kong, jimmy lai, national security